Scholarlink

Cancer remains one of the most significant health challenges globally, affecting millions in both developed and developing nations. Among various cancer types, colon cancer is the third most prevalent and the second leading cause of cancer-related fatalities, while lung cancer has the highest mortality rate. Early detection is vital for effective treatment and enhancing survival rates. Recent advancements in machine learning have facilitated the creation of computer-aided diagnostic systems, with deep learning methods greatly improving the accuracy and speed of cancer detection. These technologies offer considerable assistance to healthcare providers. This study introduces a convolutional neural network (CNN) model designed to detect and classify lung and colon cancers using histopathological images. A notable aspect of this approach is the use of Otsu segmentation for image preprocessing, which automatically identifies the optimal threshold to improve tissue differentiation and enhance cancer detection accuracy. The model was trained on a dataset comprising 25,000 histopathological images of lung and colon tissues, with 5,000 images in each class. After 50 training epochs with a batch size of 32, the model was evaluated using various performance metrics, achieving an impressive overall accuracy of 99.47%.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Learning Based Multiclass Cancer Classification: A CNN-Based Diagnostic Framework for Lung and Colon Cancer

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Saurav Mali,
  • Subrata Sinha

摘要

Cancer remains one of the most significant health challenges globally, affecting millions in both developed and developing nations. Among various cancer types, colon cancer is the third most prevalent and the second leading cause of cancer-related fatalities, while lung cancer has the highest mortality rate. Early detection is vital for effective treatment and enhancing survival rates. Recent advancements in machine learning have facilitated the creation of computer-aided diagnostic systems, with deep learning methods greatly improving the accuracy and speed of cancer detection. These technologies offer considerable assistance to healthcare providers. This study introduces a convolutional neural network (CNN) model designed to detect and classify lung and colon cancers using histopathological images. A notable aspect of this approach is the use of Otsu segmentation for image preprocessing, which automatically identifies the optimal threshold to improve tissue differentiation and enhance cancer detection accuracy. The model was trained on a dataset comprising 25,000 histopathological images of lung and colon tissues, with 5,000 images in each class. After 50 training epochs with a batch size of 32, the model was evaluated using various performance metrics, achieving an impressive overall accuracy of 99.47%.